Does FHIR need HCD?

If you work in health technology you have heard about FHIR and its promise. FHIR is a standard developed by HL7 for the electronic exchange of healthcare information. FHIR makes it easier for diverse healthcare systems such as hospitals, labs, pharmacies, and patient apps to securely and consistently share and interpret health data. The possibilities for improving the way we share and use data across systems are huge.   

The FHIR promise solves long standing user burdens on both the provider and patient level. For patients - imagine that you go to a medical office and you fill out a patient medical history form and then you never are asked to fill out it again - just review it for any changes. Killing the clipboard (1) is possible when your health information is securely shared or available at every medical encounter in a disjointed system. That means your specialty provider knows what your primary care physician has prescribed and vice versa and the burden is off the patient to ensure their medical information is up to date in every place. For providers, the promises are even bigger: FHIR can not only give you access to patient data across medical systems but it can also help you in clinical encounters by providing decision support tools, ensuring you submit your data to payers without manual effort, and helping you track measures and behaviors of your patients outside of your clinical space. When FHIR solves problems for both of these user groups, health outcomes can be improved and costs reduced. 

We know that FHIR initiatives can address some user burdens but does FHIR implementation need human-centered design (HCD)? At first glance, it may seem that implementing FHIR is an engineering problem: figure out how the systems need to talk to each other and read the implementation guide. But FHIR is already using HCD, whether it explicitly acknowledges it or not. And the possibilities for how FHIR can be leveraged to solve real world clinical problems are greater if FHIR embraces the role of HCD.

How FHIR currently uses HCD principles and techniques

Human-centered design (HCD) is a problem solving approach that centers on the needs, experiences and behaviors of people who use a tool or product. HCD typically uses a human-first approach, centering the experience of the user, approaching their problems with empathy, and getting user feedback to improve designs. If HCD has a slogan it is “Design with people, not for them.”

Despite no search results for “human-centered design” and one result for “user-centered design” (2) on the entirety of the HL7 International website, there are quite a few ways FHIR, and HL7 International, are already leveraging HCD principles.  

• The entire initiative is built on collaboration / interaction
• In describing what HL7 and its members do, the organization notes that they provide a framework (and related standards) for the exchange, integration, sharing, and retrieval of electronic health information (3). FHIR only works if collaboration and interaction are supported. 
• New FHIR specs are built with user input - in person!
• An HL7 FHIR Connectathon is centered on developing the HL7 FHIR Specification including resources, profiles, and implementation guides (IGs). The purpose of a Connectathon is to prove that the specification is complete and facilitate FHIR implementation guide maturity(4). These collaborative sessions often occur in person and could be thought of as co-creation workshops, a key component of HCD.
• FHIR benefits often cite improvements in workflows and reduced burden on the user
• In making the case for implementing FHIR, a reduction in user burden due to manual entry and submission of data if often cited (5).
• A key FHIR standard is Structured Data Capture (SDC), or a way that FHIR can help collect information in questionnaires or forms.  
• SDCs collect data from humans about humans. 

How embracing HCD can fulfill the promise of FHIR

While the presence of HCD can be felt in pockets of the FHIR ecosystem, there are many advantages of HCD that are not currently fueling the FHIR. Below I outline a few areas where HCD principles have a lot to offer FHIR and discuss how encouraging HCD collaboration in FHIR spaces could bolster current efforts and potentially unlock new benefits. 

• Measurement: What are we collecting? Are we asking the right questions?
• HCD practitioners can contribute to ensuring common SDC components are designed and configured correctly to ask questions in a way that are user friendly and collect the intended measure (e.g. ensuring validity and reliability in questionnaires). Imagine you see your doctor and are later asked to answer the question: “Do you feel your provider listened to you and adequately addressed your concerns?” This is actually two questions and any data you get from this will be essentially meaningless because you don’t know whether the provider listened, adequately addressed patient concerns, or both/neither. For patient reported measures, using HCD to improve how we ask questions can ensure the quality of the data collected.  
• Workflow: How are we collecting this information? Are we considering workflow placement?
• Another key component of HCD is empathy for users. What good is FHIR if the users find it more challenging than their established workflows? HCD can provide a lot of help here, by understanding user needs, pain points, and end-to-end journeys. This can help:
• Ensure that FHIR components that interact with users directly (e.g., pushed forms) do not present undue burden to users and fit within established workflows without additional time required. 
• Ensure that clinical providers correctly place FHIR-based data collection (such as CDS hooks) into clinical encounter workflows.
• Ensure that questionnaires are only used when they are necessary; for example, data that is typically collected as unstructured (e.g., clinical notes) would not be easily collected using FHIR and may cause users additional burden (although AI may be able to help - more on that in a later post).
• Interactions: Are we correctly considering the interactions?
• HCD typically thinks of interaction design as how a user interacts with a product, but the core principles of interaction design focus on behaviors, interactions, and outcomes. These are not so different from what should be considered for bidirectional conversations using FHIR. For example, a provider requests prior authorization from a payer before a patient procedure(6). In this example, the expected behaviors on both sides need to be considered and HCD can help define these interactions and consider the perspectives of both users (even if one or both of the users are systems). 
• Information Architecture: How is information organized and is this the best way for users to find what they need?
• In HCD we are trained to consider how information is organized and whether the user can easily find what they need. We have all visited websites and been frustrated by not being able to find something in the navigation menu that we know is on the site - this is an example of where information architecture failed. In order for FHIR to be widely implemented, users have to be able to find what they need to implement it. An example of this is the design of developer portals. These are websites where developers, business analysts, and anyone else involved in implementation can find documentation, technical specs, test data – whatever is needed to use that FHIR resource. To maximize the accessibility, reach, and usefulness of developer portals the information must be presented in a way that is findable, easy to understand, and meets all user expectations. This can be applied to other key FHIR components as well like Implementation Guides.
• Optimal User Interfaces: Can the user easily accomplish their tasks? Is the experience enjoyable?
• Speaking of developer portals, at the end of the day these are webpages and we all (should) know the value of good UI design when it comes to a pleasant and supportive user experience versus a frustrating and unsuccessful one. It also doesn’t hurt when it looks good! This can be applied to other FHIR pages, artifacts, and tools: good design can bolster the effectiveness and adoption of FHIR.
• Effective Collaboration: How can design thinking improve collaboration?
• A key component of expanding the use of FHIR are HL7 Connectathons, which rely on individuals coming together to collaborate on standards or implementation guides. HCD participation in these events can both: 1) ensure that user needs are represented; and 2) help draw out the most innovative ideas. HCD practitioners are trained in design thinking, or a way of looking at a problem, engaging different voices and perspectives, and collaboratively proposing a solution(7). Applying these techniques to Connectathons may: 1) make the end products more user friendly; 2) may help elicit more contributions from the entirety of the group; and 3) may bring more diversity in perspectives that improve the final solution. 

Risks of Not Using HCD

I hope by this point you are convinced that not only is FHIR already relying on HCD in some ways, but there are many ways that getting more HCD practitioners involved in FHIR can greatly improve what is possible for expanding how we use FHIR and the value it delivers. If you are more persuaded by the impact on the bottom line, I’ll highlight some of the risks of not including HCD when implementing FHIR:

1. Workarounds: if the implementation of FHIR doesn’t consider whether and how that fits into clinical workflows you risk providers abandoning it for a workaround, or going back to the old way of doing things. Providers don’t have a lot of time as it is - we have to make it easier, not harder.
2. Poor quality data: as I noted above, if you don’t ask the right questions or don’t phrase the question correctly you cannot make use of that data and it is a waste of time and effort to collect, clean, and analyze.
3. Lack of uptake: if your developer portal is a mess and users can’t make sense of what they find there or they cannot accomplish what they need within the UI, then they cannot use your FHIR API resource.
4. Missing puzzle piece: this might be biased, but in terms of Connectathons and the overall collaborative nature of FHIR and HL7, I can’t help but think that when there aren’t HCD practitioners present this perspective is missing and there are likely insights and suggestions that are never considered that could improve the products and adoption. (if this is your meaning).

The first three above all lead to a waste of time and resources, needing to re-do or reconsider implementation, and providing additional support, likely in the form of person-time and effort. These can be avoided by involving HCD in the design of FHIR implementation, and at additional touchpoints as needed. 

So…Now What?

There are so many ways that HCD practitioners can help make FHIR better at many levels: within projects, their organizations, and in leadership organizations like HL7. HCD practitioners need to find ways to get more involved in FHIR and need to start asking for a seat at the table. I decided to write this piece after taking the FHIR for Business Analysts course at HL7, led by my stellar colleague Benji Graham and it ignited not only my interest in learning about FHIR, but my conviction that there is a lot for HCD to contribute in this space and the time is now. FHIR is not the future of healthcare, it is the now of healthcare: 75% of EHRs in North America already incorporate a FHIR API (8).   

I encourage HCD practitioners to find a course to learn the FHIR basics or read up on your own, find out if anyone in your organization is working on FHIR implementation, and consider ways that FHIR could be leveraged for future solutions. For any non-HCD folks still reading this - ask your HCD colleagues to come to a FHIR-related meeting to get their perspective, ask them to weigh in on how you are designing your developer portal or form, or get their feedback on any assumptions you may be making about user behaviors, workflow, and preferences. As I said earlier the entire initiative is built on collaboration / interaction and it can only get better with HCD contributions. 

References:

1. https://www.cms.gov/health-tech-ecosystem/early-adopters/kill-the-clipboard
2. Searched https://www.hl7.org/search/index.cfm?q=human+centered+design on November 7, 2025.
3. https://www.hl7.org/about/FAQs/index.cfm?ref=nav#A3
4. https://www.hl7.org/events/fhir-connectathon/
5. https://ecqi.healthit.gov/fhir/about
6. See example of Da Vinci Prior Authorization: https://confluence.hl7.org/spaces/DVP/pages/345085819/Da+Vinci+Prior+Authorization+Implementers#DaVinciPriorAuthorizationImplementers-Artifacts
7. https://www.interaction-design.org/literature/topics/design-thinking?srsltid=AfmBOooRXdWmlug6wxZYfUVXBINDwbqfz3s4NCFJ8mo9t6HfgPf1luAV
8. https://codesion.com/global-fhir-adoption-statistics-a-comprehensive-overview/